Examining the expanded Agreeableness scale of the BFI-2

Nuanced HEXACO: A Meta-Analysis of HEXACO Cross-Rater Agreement, Heritability, and Rank-Order Stability

Most Five-Factor Model (FFM) questionnaire items contain unique variance that is partly heritable, stable, and consensually observable, demonstrates consistent associations with age and sex, and predicts life outcomes beyond higher order factors. Extending these findings to the HEXACO model, we meta-analyzed single-item cross-rater agreement, heritability, and 2-year stability using samples from six countries. We analyzed raw item scores and their residual variance and adjusted the estimates for measurement unreliability. The median cross-rater agreement, heritability, and stability estimates were, respectively, .30, .30, and .57, for raw items and .10, .16, and .39, for item residuals. Adjusted for reliability, the respective medians were .46 and .25 for cross-rater agreement, .46 and .39 for heritability, and .87 and .94 for stability. These results are strikingly consistent with FFM-based findings, providing nondismissible evidence that single items index a partly unique level of the trait hierarchy-personality nuances-with trait properties comparable to those of higher-order traits.

From the BFI-44 to BFI-20: Psychometric Properties of the Short Form of the Big Five Inventory

The Big Five Inventory (BFI-44) has been a useful tool for researchers for over three decades. However, the contemporary way of life has created the need for abbreviated versions of psychological instruments. We derived the number of items from the BFI-44 questionnaire in order to create a short form of the questionnaire (BFI-20). In the first study (N = 1350, 82.4% females, aged 18-60), using a range of criteria, we identified 20 items (four for each of the Big Five traits) that most optimally represent each dimension. The five-factor structure was mostly replicated in the second (N = 215, 65.1% females, aged 18-65) and third study (N = 263, 83.7% females, aged 18-42). The BFI-20 demonstrated satisfactory reliability, representativeness, homogeneity, and part-whole convergence. Despite mild attenuation, most BFI-20 correlations with schizotypy, satisfaction with life, and positive orientation remained in the same range compared to the BFI-44. The Agreeableness domain was shown to be the most challenging for capturing with four items. We discuss the advantages of our BFI-20 compared to the other two 20-item versions. In sum, we can recommend the use of this BFI-20 version as a time-efficient, satisfactory reliable, and representative questionnaire.